On This Page

📦 Installation

Get the ImpactGuard plugin running on a CS2 server in under 5 minutes.

  1. Download the Plugin

    Log into the dashboard at impactguard.xyz and download the latest plugin package.

  2. Extract Files

    Place the plugin files into the CS2 server addons directory:

    game/csgo/addons/counterstrikesharp/plugins/ImpactGuard/ ImpactGuard.dll config.json
  3. Configure

    Open config.json and set the license key (found in the dashboard).

  4. Restart the Server

    Restart the CS2 server. ImpactGuard will initialize automatically.

  5. Verify

    Check the server console for the ImpactGuard startup message, or run ig_status.

⚠️
Requirement: CounterStrikeSharp must be installed before ImpactGuard.

⚙️ Configuration

The plugin uses a single config.json file. Below is the full structure.

{ "api": { "url": "https://impactguard.xyz", "key": "YOUR_LICENSE_KEY", "timeout": 30000, "retry_attempts": 3 }, "server": { "id": "server-001", "name": "My CS2 Server", "region": "eu-west" }, "detection": { "sensitivity": 0.7, "analysis_interval": 60, "learning_enabled": true, "auto_ban": false, "ban_threshold": 0.9 }, "logging": { "level": "info", "file": "logs/impactguard.log", "console": true } }
💡
api.url should always be https://impactguard.xyz. The license key is generated in the dashboard.

📋 Key Parameters

The most important settings at a glance.

api.key string

Required
License key generated from the ImpactGuard dashboard at impactguard.xyz.

detection.sensitivity float

Default: 0.7  (range 0.1 – 1.0)
Higher values catch more cheaters but increase false positives. Start at 0.7.

detection.auto_ban boolean

Default: false
Automatically ban flagged players. Keep disabled until the system has calibrated (~100 samples).

detection.ban_threshold float

Default: 0.9
Minimum confidence score required for an automatic ban when auto_ban is enabled.

detection.learning_enabled boolean

Default: true
Enables the adaptive ML engine. Needs at least 100 player samples to calibrate.

logging.level string

Default: "info"
Options: debug, info, warn, error. Use debug when diagnosing issues.

🛡️ Detection Features

ImpactGuard runs 28 specialized detection modules across every cheat category.

Aimbot Detection
Wallhack Detection
Movement Analysis
Mouse Pattern Analysis
Reaction Time Analysis
Game Behavior Analysis
ML Pattern Recognition
Statistical Anomaly Detection
ModuleWhat It DetectsAccuracy
Aimbot DetectionDetects aim assistance through lock-on speed, angle snapping, and target acquisition patterns95%
Wallhack DetectionIdentifies players tracking enemies through walls via crosshair placement and pre-aim behavior93%
Movement AnalysisFlags unnatural positioning, bunnyhop scripts, and impossible strafing patterns88%
Mouse Pattern AnalysisAnalyzes mouse trajectories, micro-adjustments, and humanness of aim curves94%
Reaction Time AnalysisDetects superhuman response times to visual stimuli and audio cues91%
Game Behavior AnalysisEvaluates overall gameplay patterns including economy decisions and rotation timing96%
ML Pattern RecognitionIsolationForest model scoring behavioral anomalies against learned baselines97%
Statistical Anomaly DetectionFlags statistically impossible performance metrics like headshot ratio and KD spikes92%
💡
Note: Accuracy improves over time as the ML system collects more local data from the server.

💻 Console Commands

Admin commands available in the CS2 server console.

CommandDescription
ig_statusShow ImpactGuard status and connection info
ig_reloadReload config.json without restarting
ig_analyze [player]Manually trigger analysis for a player
ig_ban [player] [reason]Manually ban a player
ig_unban [steam_id]Unban a player by Steam ID
ig_configDisplay the current running configuration

🧠 Learning System

ImpactGuard uses an adaptive ML engine that gets more accurate the longer it runs.

How It Works

Every player action flows through a four-stage pipeline: Data Collection → Feature Extraction → Pattern Analysis → Decision. The IsolationForest model scores behavioral anomalies and produces a confidence-based risk score that triggers flags, kicks, or bans.

30GB
Max Storage
24h
Auto Cleanup
7d
Training Cycle
100
Min Samples

Cross-Server Intelligence

When enabled, the plugin anonymously shares behavioral patterns with the ImpactGuard network. This collective intelligence strengthens detection across all participating servers. Player identities are hashed — no personal data is ever transmitted. This can be disabled at any time in config.

Expected results: After ~100 samples and one training cycle, most servers see 95%+ accuracy with under 1% false positives.

📊 Web Dashboard

The ImpactGuard dashboard at impactguard.xyz provides a full management interface.

Live Monitoring

Real-time player analysis, active flags, and server status at a glance.

Ban Management

Review flagged players, confirm or dismiss detections, and manage ban lists.

Detection Analytics

Charts and statistics showing detection rates, false positive trends, and module performance.

License Management

Generate, rotate, and revoke license keys for the plugin.

🔧 Troubleshooting

Common issues and how to resolve them.

Plugin Not Loading

API Connection Failed

High False Positives

Low Detection Accuracy

⚠️
Below 85% accuracy? The ML model needs at least 100 player samples to calibrate properly. New servers should expect lower accuracy for the first few days.

🔒 Security

Important notes about keeping the plugin and server protected.

⚠️
Never share the license key. Treat it like a password. If it's exposed, rotate it immediately from the dashboard.